Welcome to the latest installment of “Shocked Director Discovers the Internet is Trash,” featuring Valerie Veatch and her brave journey into the heart of OpenAI’s Sora. In a move that truly redefines the word “hyperbole,” Veatch has looked at a text-to-video generator and decided it’s not just a tool for making uncanny valley fever dreams—it’s actually a refreshing glass of eugenics-flavored Kool-Aid.

First, let’s address the “shock” that a generative AI model, trained on the vast, unfiltered dumpster fire of human history known as the internet, might output something biased. Identifying racism and sexism in a large language model is like identifying salt in the ocean. It’s not a discovery; it’s a prerequisite. To be “shocked” that an algorithm reflects human prejudice is to admit you’ve never spent more than five minutes on a social media comment thread or, heaven forbid, looked at the dataset it was fed. It’s a stochastic mirror, Valerie. If you don’t like what you see, maybe stop blaming the glass and look at the source material: us.

Then there’s the claim that the AI-enthusiast community “does not seem to care.” This is a classic case of confusing “trying to figure out how to make a cat fly a plane” with “orchestrating a global movement for selective breeding.” Most people using Sora are trying to get the prompt engineering right so the characters don’t have seventeen fingers and three heads. To suggest that a community of digital artists is indifferent to social justice because they aren’t performing a 48-hour vigil over every pixelated artifact is a reach so long it’s practically atmospheric.

The comparison to eugenics is where the logic really enters orbit. Eugenics is a pseudo-scientific movement aimed at improving the human race through controlled breeding and, historically, horrific state-sponsored violence. Sora is a software program that lets you turn “Cyberpunk Corgi” into a 10-second MP4. The leap from “this AI generated a stereotypical image” to “this is the digital reincarnation of Francis Galton” is the kind of intellectual parkour that only thrives in a specific type of tech-criticism echo chamber.

The assumption here is that if a technology isn’t born perfectly virtuous and sanitized of all human failing, it is inherently evil. This “all-or-nothing” moral gatekeeping ignores the reality of how technology evolves. We didn’t ban the printing press because people printed hate speech; we didn’t delete the internet because of pop-up ads for miracle cures. We iterated. But apparently, for the “gen AI is eugenics” crowd, if the tool isn’t a moral philosopher out of the box, it’s a weapon of mass destruction.

Let’s be real: AI bias is a legitimate engineering challenge and a social concern. It requires rigorous auditing, better datasets, and constant refinement. But calling it “eugenics” doesn’t help solve the problem—it just wins you points in a very specific, very loud corner of the internet. It turns a complex technical and sociological issue into a catchy, inflammatory headline.

So, for those of us still drinking the “Kool-Aid,” we’ll be over here trying to figure out how to get the AI to render a realistic-looking strawberry without it turning into a metaphor for the collapse of Western civilization. It’s a tough job, but someone has to do it.